1

Power Asymmetries and the Origin of Constitutions: Latin America,

1987-2009

Javier Corrales

Amherst College

[email protected]

June 2009

12,200 words 2

Abstract

This article explores the origins of constitutions, and within them, levels of power concentration in the Executive branch. By looking at the 10 cases of constitutional rewrite in Latin America and several cases of aborted change since the 1980s, I develop a theory of constitutional change based on the notion of power asymmetries, defined as the power differential between incumbents and opposition forces at the negotiations. Under conditions of reduced power asymmetry, i.e., when both incumbents and opposition forces enjoy comparable levels of power, constituent assemblies are more likely yield constitutions that curtail presidential powers relative to the status quo. When the opposition is weak, assemblies will instead expand presidential powers. When the incumbent is weak, he or she will abort the process of constitutional rewrite. I conclude with a brief discussion of how this finding challenges some structuralist prescriptions for democratic renewal.

3

This article addresses two questions that are central to comparativists in general and

Latin Americanists in particular. First, what are the origins of institutions, and specifically, constitutions? Significant research exists on the effects of different types of institutions, even constitutions,1 but less is known about their origins. 2

Second, what are the origins of different forms of presidential powers in new constitutions? Research has shown that democratic constitutions vary according to how much power they grant to the executive branch.3 But less is known about the conditions that give rise to such variation.

This article examines these two questions by looking at the experience of Latin America.

The region is known worldwide as the land of presidentialism. It should also be known as the land of constitutional rewrite. Since the transition to democracy in the late 1970s, the region has seen ten cases of major constitutional rewrite:4 Nicaragua (1987), Brazil (1988), Paraguay

1 For example, Reynolds 2002; Persson and Tabellini 2003.

2 Research on the origins of constitutions is in its infancy. For Latin America, the best papers are

Negretto (2009), Nolte (2009) and Hartlyn and Luna (2009).

3 Shugart and Carey 1992.

4 Constitutional change occurs through two mechanisms: complete rewrites (replacement) and amendments. This paper looks at rewrites only, although change through amendment has been significant. Nolte, for instance, counts 312 amendments in Latin America between 1978-

2008. This paper will not look at the amendment process since it “responds to a different logic:” amendments are narrower in scope, are approved by a smaller number of actors 4

(1991), Colombia (1991), Argentina (1994), Peru (1995), Ecuador (1997), Venezuela (1999),

Ecuador (2008), and Bolivia (2009). 5 All these post-transition rewrites occurred through highly participatory processes, that is, through constituent assemblies composed of elected delegates, rather than by negotiations among a restricted circle of elites. 6

None of these assemblies abolished presidentialism. But the resulting documents changed presidential powers in different directions. Some assemblies expanded presidential powers relative to the status quo; others hardly changed, while still others reduced presidential powers. These cases thus provide an opportunity to study the factors that give rise to constitutions, and within each constitution, levels of power concentration, while controlling for historical era and region.

I will argue that this variation in presidential powers is the result of power asymmetries between incumbent ( I) and opposition ( O) forces. I define power asymmetry as the power differential between these actors. At the most basic level, cases can display three forms of asymmetry: I and O can have comparable levels of power (reduced asymmetry); I can be far stronger than O (large pro-I asymmetry); or O can be far stronger than I (negative, or pro-O

(usually, the Executive branch in conjunction with a few parliamentarians, and tend to be mostly about policy- related matters (see Nolte 2009:8; see also Hartlyn and Luna 2009).

5 The 1983 El Salvador and 1985 Guatemala constitutional rewrites are not included because they were enacted under authoritarian settings (a non-elected government).

6 On the different ways in which constitutions can change, especially the distinction among legislature-based, constituent assembly-based, and president-based changes, see Elkins,

Ginsburg and Blount 2009. 5 asymmetry). My argument is that conditions of reduced asymmetry yield constitutions that lessen presidential powers. Large asymmetry, in contrast, yields constitutions that expand presidential powers, and negative asymmetry increases the chances of I aborting constitutional rewrites. These propositions are based on the premise that, in the politics of constitutional change, Is prefer expanding presidential powers, while Os prefer instead to expand the powers of non-Executive branches. In short, power asymmetries determine each actor’s bargaining power, and thus, the incidence and content of constitutions.

I. Constitutions as Pacts

A fruitful starting point for generating a theory about the origins and content of constitutions is to draw from the literature on democratic pacts, which flourished in the late

1980s when transitology, or the study of transitions to democracy, was in vogue. The literature defines pacts as any agreement or compromise among political forces. Pacts can be as simple as a document of understanding about procedures or policies, or they can be as encompassing as a national constitution. A constitution, in turn, can be defined as the “codes of norms which aspire to regulate the allocation of powers, functions, and duties among the various agencies and officers of government, and to define the relationship between these and the public. Like pacts, constitutions typically emerge as a result of negotiations among actors with different bargaining power.

Also like pacts, democratic constitutions are quintessential functional institutions: actors draft and sign them hoping restrain their counterparts and mitigate their own political 6 insecurity. Because pacts are meant to serve as antidotes to polarization and mutual suspicion, they help pave the way for democratization.7 They are also a mechanism through which actors offer concessions to their opponents in return for some gains.8

What determines whether these constitutional pacts will emerge, and whether they will in fact offer solutions to political insecurity? The theoretical challenge is to specify the conditions that generate the different strategies that actors adopt when making pacts,9 or for our purposes, constitutions. Constitutions can take multiple forms and occur in very diverse political settings. Yet, there is one simple distinction in pact-making that shape the key contours of the outcome: the balance of forces among signatories. If the balance of power among signatories is deeply asymmetrical, the constitution either does not get rewritten, or if changed, becomes intensely biased toward I’s preference.

II. Pacts as Bargaining Outcomes

Arguments about successful pact bargaining are predicated on four basic claims. First, successful democratic pact-making, in fact, all forms of rights-granting, depend on the strategic interaction between two sides of elites—namely, Is and Os.10 Second, pact-making depends on

7 O'Donnell and Schmitter 1986; Di Palma 1990; Karl 1990, 1-23; Karl and Schmitter 1991, 269-

84; Przeworski 1991; Reynolds 2002.

8 Acemoglu and Robinson 2006.

9 See Snyder and Mahoney 1999.

10 Przeworski 1991; Higley and Gunther 1992; Tilly 1992; Negretto 1998; Tilly 1998. 7 the prevalence of soft-liners on each side.11 Soft-liners are actors who prefer to deal with opponents through peaceful, rule-bound competition rather than through open confrontation or noncooperation.12 Third, successful pact-making depends on “mutual guarantees.” Each actor must offer guarantees that it will not threaten the “vital interests” of counterparts.13 And finally, as a corollary to the previous point, the most important guarantee that must be offered is “credible bounds on the behavior of political officials.”14

Yet, this picture, however rich, is theoretically incomplete. First, except for the proposition that soft-line strategies must prevail, these arguments are mostly prescriptive, outlining the contours of a successful pact, rather than the conditions that lead to their emergence.

Second, the prevalence of soft-liners is not enough to explain the propensity of pacts to yield condition four; namely, lowering the stakes of holding office. There is no question that democratic pact-making requires Is to adopt a soft-line position: it must agree to offer concessions to opponents and, more crucially, accept rules of self-restraint. But I will only agree to rules of self-restraint if it feels sufficiently pressured to do so. As Kantor 15 argued decades ago in his study of Latin American constitutions, the way to limit the power of the president is

11 O'Donnell and Schmitter 1986.

12 Linz 1978.

13 Karl 1990; Karl and Schmitter 1991.

14 Schedler, Plattner et al. 1999; North, Summerhill et al. 2000, 23-29; Weingast 2004.

15 Kantor 1977. 8 to create “competing centres of power.” For those power centers to emerge, O must be relatively strong (an institutional condition) and willing to exert pressure (a strategic condition).

O therefore cannot be as moderate as I.16 If O doesn’t pressure I sufficiently, I will not have an incentive to yield.17 Essentially, I’s propensity to be moderate may depend on O’s propensity to pressure.

Yet, as O’Donnell and Schmitter emphasized, O must not be excessive either. If O becomes too threatening, it will scare I and thus ruin the chance of a negotiation. Thus, when

Acemoglu and Robinson 18 talk about democracy arising through revolutionary means, they may be going overboard. In the face of hard-core confrontation, or even insurgency, rulers are more

(or equally) likely to turn to guns in self-defense than to pact-making; and if they cannot repress

(because the costs of repressing are high), they will at least seek to split the opposition: negotiate with some and repress the others, which is not a fully democratic result.

O’s central dilemma in the politics of democratic pact-making is therefore that it must offer a combination of some hard-line resistance (to force I to concede) and also some degree of moderation (to offer incentives for I to negotiate). It cannot be one or the other: it must be both, as hard as that may sound.

16 Bermeo 1997.

17 Tilly 1992; Tilly 1998.

18 Acemoglu and Robinson 2006. 9

For these reason, high pro-I asymmetry is fatal for a balanced outcome.19 With few seats, O will lack the necessary bargaining leverage to extract favorable concessions from I. O needs bargaining leverage because it has a particularly ambitious agenda, perhaps more so than Is.20 O needs the constitution to offer not just protection (or minimum guarantees), but also provisions that make it easier for it to leave its opposition status.21 This is the paradox of constitution-making: O needs the most out of the pact; yet, it often faces the risk of being the weaker party in the negotiation.

The point of the discussion thus far is to demonstrate that the balance of forces between I and O helps to explain both the pact and whether the pact will offer new opportunities and new restraints on I. Following Kantor,22 Geddes,23 Colomer,24 Bermeo,25

Negretto,26 Hartlyn and Luna, 27 I posit that under conditions of reduced asymmetry (e.g., where

I and O are both strong), constitutional negotiations will impose limits on I’s preferences. As

19 Arato 2005, 17.

20 Horowitz 2002, 27.

21 Lijphart 1991,72-84; Lijphart 1992.

22 Kantor 1977.

23 Geddes 1994.

24 Colomer 1995, 74-85.

25 Bermeo 1997, 305-22.

26 Negretto 1998; 2001; 2002.

27 Hartlyn and Luna 2009. 10

Olson argues, democracy emerges when there is a “broadly equal dispersion of power that makes it imprudent for any leader or group to attempt to overpower the other.” 28 Under large asymmetry, on the other hand, I has no incentive to offer guarantees to O because it does not fear it, and O is not strong enough to compel I to yield.

III. A Bargaining Model with Power Asymmetries

Figures 1, 2 and 3 offer a simple diagram of my argument. In approaching the prospects of constitutional rewrite, both I and O want to maximize their respective formal powers. Thus, I prefers to move the status quo (SQ f) northwardly toward ideal point x I, and O prefers to move easterly toward ideal point x O (see Figure 1). P I represents I’s preference frontier; points to the right of P I will be rejected by I. Likewise, P O represents O’s preference frontier: points to the left will be rejected by O. The shaded petal represents the agreement zone, or win-set.

[Figure 1]

Figure 1 does not capture each actor’s bargaining power, and thus, whether the negotiations will fall inside or outside the shaded area. My contention is that power asymmetry determines bargaining power. Figure 1 can be adapted to represent the status quo, not in terms of formal powers (as specified by how the existing constitution distributes formal powers), but in terms of power asymmetry: each actor’s share of votes/seats in a given election/institution. Figure 2 captures this adaptation. Point SQ PA in Figure 2 represents a situation of reduced asymmetry between I and O. Since by definition each actor has

28 Olson 2000, 31. 11 comparable bargaining power, each will be able to restrain the other. Thus, the negotiations will move diagonally into the shaded area (rather than northwardly or easterly). Figure 2 thus shows that reduced asymmetry will yield an agreement in which both I and O will win and lose some, and neither will win significantly more than the other.

[Figure 2]

Figure 3 represents large, pro-I power asymmetry. I enjoys more bargaining power to push the status quo in its desired direction, increasing the chance of moving outside the common agreement space. I will win far more than O, and O will find the outcome displeasing.

Likewise, under negative asymmetry (not diagrammed), O is able to push the negotiations outside the shaded area, but into I’s rejection zone. In that case, I will have every incentive, and probably sufficient formal powers, to abort the process of pact-making.

[Figure 3]

IV. Pivotal issues: Taming or Expanding Presidential Powers

I have discussed the actors and factors that determine who wins in pact-making, but not the issues over which they fight. The politics of constitutional change engages a multitude of issues that vary across countries and times. However, one issue seems common to all democracies: how much to tame or expand presidential powers.

Armony and Schamis argue that all majority-based democracies are susceptible to presidents who “strive for greater autonomy…circumvent congressional and judicial 12 oversight…and ultimately thwart[ing] the principle of separation of powers.”29 If all democracies face a “perpetual tension” between the impulse of executives to expand their powers, and the countervailing tendency of groups (in office or in society) to control and oversee the Executive branch, these tensions are even more conspicuous during constitutional processes. As Elkins, Ginsburg and Blount argue, actors will want to maximize the power of the office they hold (what they call “self-dealing). 30 This means that pro-I forces typically push for presidential powers, what Hartlyn and Luna (2009) refer to as “power concentration,” while opponents seek power diffusion.

In Latin America, the sectors of Os that most strongly resisted power concentration (on the presidency) were the organized political parties in the opposition. They were cognizant of the fact that hyperpresidentialism, in the short term, hurts opposition parties the most. They thus approached the politics of constitution-making with a strong preference on behalf of lessening power concentration and expanding instead the powers of non-Executive branch institutions (where they have a presence).

V. Hypotheses and Case Justification

29 See Armony and Schamis 2005, 116-117.

30 On how I’s demand for more presidential powers rose in Latin America since the 1980s, see

O’Donnell’s (1994) famous essay on “delegative democracy,” which is both a summary and a normative lamentation of this heightened demand.

13

Now that we have specified the different actors (I and O) and factors (power asymmetries) that shape constitutional contents, as well as a pivotal issue of negotiations

(presidential powers), we can formulate a set of hypotheses:

H1: Under conditions of expanded asymmetry, assemblies will yield constitutions that

increase power concentration (i.e., expansion of presidential powers).

H2: Under conditions of reduced asymmetry, assemblies will yield constitutions that

diffuse power (reduction in presidential powers), if the status quo constitution was too

biased toward Executive powers.

H3: Under conditions of reduced asymmetry, assemblies will not reduce presidential

powers significantly, if the status quo constitution was not too biased toward power

concentration.

There is one more issue to consider: there are two measurements of power asymmetry that matter. The first is what could be called “institutional asymmetry”: the distance between I and O within any given political institution such as Congress, governorships, party votes. 31 The

31 Amorim Neto (2002) applies a similar argument to Executive-legislative relations under fragmented political party systems, what he calls “cabinet coalescence.” 14 other is what could be called “table asymmetry”: the distance between I and O at the negotiating table itself, measured by share of seats held by each.

In some cases, institutional and table asymmetries will be dealigned: the power differential at the negotiating table might differ from the power differential throughout the political system as a whole. For instance, O can end up with more power (shares of seats) in the constituent assembly than it obtained in the most recent electoral contest or in Congress, and vice versa. In the former type of dealignment, the resulting constitution will yield more victories for O (H2 and H3). In the latter, the resulting constitution will yield not only far more victories for I (H1), but also increase the probability political instability in the short term. O will be insufficiently powerful to stop I from prevailing at the negotiations (due to low table asymmetry), but sufficiently powerful to disrupt implementation (due to strong pro-O institutional asymmetry). Thus, I offer the following “dealignment” hypothesis:

H4: If table asymmetry is negative (pro-O) and significantly de-aligned with institutional

asymmetry, the assembly will yield a constitution that increases power concentration

and the chance of political unrest, at least in the short term.

I will examine these hypotheses by looking at the 10 Latin American cases of constitutional change through constituent assemblies. Looking at 10 cases offers some advantages over the more traditional single-case study or large-n studies of constitutional change. The merit of single case studies is that the researcher can draw from a variety of 15 authoritative sources to understand subtleties and trace causal and effect mechanisms. The disadvantage is uncertainty about generalizability across other cases. The virtue of large-n studies is, of course, the opposite. Propositions can be generalized because they can be tested across cases and time spans. However, large-n studies have the disadvantage of being somewhat blind to the context of constitutions, that is, to pivotal issues, and how those issues get resolved, which is key to whether constitutions become acceptable or not. Furthermore, many large-n studies of constitutional change do not test situations of aborted constitutions, and thus, they can never offer a complete theory about the rise of constitutional change.

By focusing on multiple cases in exclusively one region (Latin America) during only one historical era, I can examine causal arguments while simultaneously controlling for factors known to affect constitutions such as regional variables, historical eras, and cultural heritage. I can also be more sensitive to pivotal issues, in this case, the presidentialism, and thus understand better the short-term response of actors. And as I will show at the end of the article, my method will allow me to study instances of aborted change.

VI. Measuring Presidential Powers

My first task is to rank the degree of presidential powers across cases. I developed an index that expands on Shugart and Carey’s 32 seminal work on presidential powers. Shugart and

Carey organize their scores of presidential powers into two broad categories: legislative powers and non-legislative powers. Since then, other authors have identified additional

32 Shugart and Carey 1992. 16 presidential powers, within and outside those categories. My index draws from these authors, especially Hartlyn and Luna,33 Lutz,34 Alberts,35 and Bejarano.36 In addition, I added my own categories, based on my reading of key items that emerged in the new Latin American constitutions (see Appendix 1). Country codings are available from the author.

The scorecard produced a raw score for each constitution, which I then converted into an index ranging from 0 (no presidential powers) to 1.0 (full presidential powers). In a democracy, values approaching 0 and 1 are untenable, by definition: all democracies rely on an executive branch, which makes 0 impossible, and on limited government, which makes 1 also impossible. Thus, in my index of presidential powers, democracies are expected to vary within a narrow range that never comes close to either 0 or 1.

Graph 1 shows the pre and after scores for each Latin American case of constitutional rewrite. The outcomes vary starkly. But before discussing this variation, I would like to offer a word about the index’s reliability. The index is very clear about two of the most studied cases of constitutional change in the 1990s: Alberto Fujimori’s constitution in Peru (1993), which expanded presidential powers, and César Gaviria’s constitution in Colombia (1991), which curtailed presidential powers. This result is perfectly consistent with what the qualitative literature on these cases has concluded. That my index reflects well what the qualitative literature documents adds confidence on its reliability.

33 Hartlyn and Luna 2009.

34 Lutz 2006.

35 Alberts 2006.

36 Bejarano 2006. 17

[Graph 1]

Nevertheless, this index is not perfect, on at least two counts. First, it measures presidential powers mostly in terms of the relationship between the Executive Branch and the other branches of government, while ignoring other dimensions, such as the electoral system and party regulations, and offering rather basic measurements of the state’s role in the economy and variations in subnational powers. Second, the index assigns equal weight to every category, when in reality, some categories may be far more important than others in general, or across countries, or across time. For this reason, it is better to use the index less as a way to rank countries among each other, and more as a way to gauge in change within countries, which is exactly how I propose to use the index.

VII. Hypotheses 1, 2 and 3

Overall, the index reveals three clusters of outcomes (Graph 1): Peru (1993), Venezuela

(1999), and Ecuador (2008) expanded presidential powers. Nicaragua (1987), Paraguay (1992),

Ecuador (1998) and Colombia (1991) reduced presidential powers. Argentina and Brazil show small movements, while Bolivia goes back and forth.

[Graph 1]

To what extent do power asymmetries explain these outcomes? Table 1 shows one measure of institutional power asymmetry for each case: electoral results in the most recent election prior to the formation of the constituent assembly. It also shows table asymmetries: distribution of seats at the negotiating body. Despite the variety of triggering circumstances, most Is confronted (to their surprise) significantly reduced table asymmetries at the constituent 18 assemblies. This was true even in Argentina, Peru, and Bolivia where presidents Carlos

Menem, Fujimori, and Evo Morales expected large table asymmetries given their high approval ratings in public opinion polls 37 and yet received constituent assemblies that had less than favorable table asymmetries. The sole exceptions were Venezuela 1999 and Ecuador 2008, where table asymmetries were greater than I anticipated or enjoyed previously.

Table 1 shows the hypotheses confirmed by each case. Negative or reduced table asymmetries are associated with reduced presidentialism: Nicaragua, Colombia, Ecuador 08,

Argentina, and Bolivia 09. Large table asymmetries are associated with expanded presidential powers: Ecuador 08, Peru, Venezuela, and Bolivia 08.

[Table 1]

Graph 2 illustrates this relationship more clearly. It compares table asymmetries

(vertical axis) with percent change in the index between the the status quo ante and the new constitution. As table asymmetry moves from negative to positive values, presidential powers rise.

[Graph 2]

Paraguay and Nicaragua specifically support H2. The status quo ante had excessive power concentration (constitutions from the authoritarian era) and table asymmetry was far lower than institutional asymmetry in both cases. The result was a reduction in presidential powers, as is expected of cases undergoing transitions to democracy.

In all cases of reduced table asymmetry, O forces obtained crucial victories, including curtailment of Executive branch powers, empowerment of non-presidential institutions,

37 See Kay 1996; Calvert 2002; Gamarra 2007, respectively. 19 deepening of decentralization. Small and nontraditional forces obtained more access to political resources such as state funding. In Argentina and Peru, presidents obtained one of their most significant preference (consecutive re-election, expanded decree authority, and in

Peru, a smaller, unicameral congress), but they had to concede to O more than they had planned.38

In Argentina mutual victories were clearer, confirming H3. As Calvert 39 argues, Menem

“had to pay a high price” for his re-election concession: a reduction in the presidential mandate from six to four years, delegation of some presidential powers to a Chief of Cabinet who can be removed by the legislature, a runoff election, the creation of an autonomous government in Buenos Aires, an increase in the number of senators. In addition, the new constitution regulates the president’s leeway to regulate by decree.

The one seemingly inconsistent case is Brazil 88: despite reduced asymmetry, the constitution retained power concentration,40 which seems to challenge H2. Yet, seen from the point of view of realized intentions, the case is still consistent with my argument. Despite his broad popularity, the incumbent José Sarney was unable to increase presidential powers as he had intended. Sarney was anxious for greater powers because he was engaged in a losing battle against runaway inflation. With two severe adjustment plans failing,41 Sarney argued for greater powers to control fiscal spending. Instead, the 1988 constitution increased the

38 Jones 1997, 290-98; Schmidt 2000; Llanos 2003, 37-42.

39 Calvert 2002.

40 Alston, et al. 2005.

41 Plan Cruzado of 1986 and Plan Bresser of 1987. 20 national state’s financial commitments to various sectors of society and transferred portions of federal tax revenues to subnational governments, without requiring them to provide additional services.

VI. Hypothesis 4: Dealignment

The issue of dealignment between table and institutional asymmetries (H4) deserves further attention. In some cases, table asymmetries reflected closely the country’s general distribution of power between I and O. But in other cases, there was dealignment in favor or either O or I, each with different outcomes in terms of content and effects of constitutional change.

A. Pro-O De-alignment: Colombia 1991 and Nicaragua 1987

There are two clear cases of de-alignment in favor of O, each occurring under Is with divergent ideological proclivities. The first is Nicaragua 1987. Here, constitutional change occurred under the auspices of a revolutionary-leftist I, president Daniel Ortega of the Frente

Sandinista de Liberación Nacional (FSLN). In the 1984 legislative elections—billed as the preamble to a constituent assembly— the FSLN won a comfortable power advantage (see Table

1). Yet, the FSLN decided to carry out the constitutional rewrite through a political subsidy:

The FSLN-dominated legislature appointed a commission composed of 12 members from the

FSLN and 10 members from other parties. Essentially, I agreed to a smaller proportion of seats 21 than it held in the legislature,42 thereby reducing asymmetry from 26.5 in the legislature to 9.0 at the negotiating table. 43 While right-wing parties were excluded, 44 the key point is that table asymmetry became dealigned from institutional asymmetry in the direction of O.

The result was a reduction in presidential powers. This is clear in comparison to two benchmarks. The first is the last Somoza constitution of 1974 (see Graph 1), which, although

“bogus” 45 because Somoza ruled by fiat, did contain some “advanced elements,”46 including separation of power, a prohibition on re-election, and a wide array of social rights. A second benchmark is the “Fundamental Statutes of the Republic,” the short document that served as the de facto Sandinista constitution between 1979 and 1987 and which justified FSLN rule by fiat. 47 Relative to either benchmark, the 1987 Constitution reduced presidential powers and included “liberal” elements such as four separate branches of government and the multiparty system.48 The resulting power diffusion partly explains why so many O delegates (74 percent) agreed to sign the 1987 Constitution and why the ruling party had to resort to govern by decree after the signing (it felt in need of far more formal powers than the constitution granted in

42 Reding 1987.

43 Jonas and Stein 1990.

44 United States Department of State 1991 (1987).

45 Mijeski 1991.

46 Escovar Fornos 2000, 134.

47 Azicri 1991; Prevost 1997.

48 McConnell 1997. 22 order to advance its revolutionary agenda), and why O was able to capture the presidency in

1990—previously unavailable channels of contestation were opened.

The second case of dealignment is Colombia 1991. This time, the subsidy to O came under Gaviria, a president who was Ortega’s ideological opposite, i.e., more committed to a liberal, market democracy. Yet, like Ortega, Gaviria approved a selection rule for electing delegates to the constituent assembly that was very favorable to O, yielding a dramatic reversal from pro-I institutional asymmetry to a pro-O table symmetry, with no single political bloc dominating.

Consequently, the 1991 Colombian constitution also curtailed presidential powers. It introduced the election of governors, gave governors more resources, reduced the presidential powers in many policy areas, limited the president’s decree powers, and increased the power of the legislature to overrule presidential veto.49

These pro-O political “subsidies” in Nicaragua and Colombia were the result of pressure.

In Nicaragua, sectors of O were well armed (the Contras), and foreign actors (the United States, the Contadora group) pressed I for accommodation.50 In Colombia, the pressure came mostly from within. As in Nicaragua, a sector of Colombia’s O was armed (the M-19), but also, the non-armed sectors (especially the media and the student movement) pressured hard for a pro-O election rule.51 leading to reduced table asymmetry, and subsequently, to reduction in

49 Archer and Shugart 1997; Cárdenas, Junguito et al. 2004.

50 See Pastor 2002.

51 Van Cott 2000, 54-60. 23 presidential powers. This confirm the points that a mix of hard and soft-line opposition was crucial for abating presidential powers.

B. Pro-I Dealignment: Venezuela 1999 and Bolivia 2007-09

In contrast, the most extreme case of pro-I dealignment is Venezuela in 1999, which explains why Venezuela 1999 is such a regional outlier in terms of expansion of presidential powers. At inauguration time in early 1999, Hugo Chávez was overwhelmingly popular, but he faced negative asymmetry in the legislature. His popularity encouraged him to pursue constitutional rewrite; his negative asymmetry in Congress encouraged him to avoid the congressional route, as mandated by the existing constitution. He proposed a constituent assembly with a selection rule that was deliberately designed to minimize the electoral success of O forces. The result was a constituent assembly that had the most extreme, pro-I table asymmetry in the history of democratic Venezuela and Latin America—88.6 points. 52

The constitutional process was thus a pro-I affair from the start. The president obtained almost every one of his wishes. By September 1999, the Constituent Assembly, with the help of

Chavista supporters in the streets, deactivated congress and penalized federal judges who expressed reservations about the Assembly’s extraordinary powers. The president of the

52 Although the total votes obtained by opposition and independent candidates reached approximately 30 percent, only six made it into the Constitutional Assembly. For an explanation of how the electoral system and electoral strategies led to this underrepresentation, see Penfold 1999, Maingón, Pérez Baralt and Sonntag 2000; Crisp and

Johnson 2001; and Coppedge 2003. 24

Supreme Court, Cecilia Sosa, resigned, arguing that the courts preferred to “commit suicide” than to suffer assassination.53 In November, the directorate of the Assembly approved a decree restricting the extent of the debate, in order to rush the signing of the constitutions. In

December, a week after the electorate approved the new Constitution, the Constituent

Assembly decreed the “Public Power Transition Regime,” whereby the existing Congress, the state legislatures, the authorities in the Supreme Court of Justice, Attorney General, the

National Comptroller, and the National Electoral Council were disbanded. The Assembly proceeded to appoint “provisional authorities” for these posts, including a twenty-one-person provisional legislature (the Congresillo ). These acts shared the same logic—abolishing all institutions featuring reduced asymmetry, all done in a completely “closed process,” to use the phrase preferred by Martha Harnecker, one Chávez’s most renowned supporters.54

The result of extreme table asymmetry in 1999 was the largest expansion of presidential powers in both relative and absolute terms . No doubt, this constitution conferred new rights on citizens: the constitution is written in gender-inclusive language, creates the mechanism of popular referenda, and raises traditional social rights (e.g., to education and health) and less traditional rights (e.g., environmental protection, motherhood, indigenous rights) to the same levels as political rights.55 Furthermore, the president did not obtain the power to rule by decree in some areas, as in Argentina, Brazil, Ecuador, Paraguay and Peru. These limitations on presidential powers are important because they mean that presidents still required “partisan

53 Combellas 2003.

54 See Chávez, Harnecker et al. 2005, 48-50.

55 Wilpert 2003. 25 powers” (Congressional majorities, disciplined ruling party) to rule unencumbered. However, there is no question that the biggest winner—in terms of net powers—was the Executive branch.56

The expansion of formal presidential powers in the context of de-aligned table and institutional asymmetries had predictable consequences. First, I used the newly acquired formal powers to expand its powers further and attack the opposition not long after the approval of the new constitution. Second, O was able to mobilize against these encroachments: between 2001 and 2004, Venezuela plunged into the highest degree of political instability in its democratic history, displaying the most massive political protests ever seen in South America.

Venezuela’s instability subsided only after a series of events lowered O’s institutional power. These events included undemocratic moves by I (firing 20,000 opponents from the state-oil company in 2003, launching social programs to benefit mostly loyalists, undermining the independence of the court and electoral bodies; discriminating against job applicants who signed petitions to hold a recall referendum), mistakes by O (the 2005 decision to boycott legislative elections, which reduced to zero O’s representation in the National Assembly), and non-transparent electoral processes (the 2004 recall referendum, in which I won in opaque circumstances).57

56 For qualitative assessments reaching the same conclusions, see Márquez Ferrer 1998; Alvarez

Díaz 1999; Revista SIC 1999, 434-35; Virtuoso 1999; Maingón, Pérez Baralt, and Sonntag, 2000;

Blanco 2003, 250-54.

57 See Corrales and Penfold 2007. 26

A second case of pro-I de-alignment, though short-lived, was Bolivia 07-09. More clearly than in Venezuela, Bolivia had reduced institutional asymmetries at the start of the process. I was majoritarian in some regions: La Paz, Oruro, Potosí, Chuquisaca, whereas O was majoritarian in the other regions: Santa Cruz, Beni, Tarija, and Pando.58 Furthermore, table asymmetry at the Constituent assembly was a mere 7.4, with I failing to replicate the overwhelming majority that Chávez gained in 1999. In addition, rather than fragment as in

Venezuela, O forces in Bolivia coalesced into a more unified political party, PODEMOS. In short

I was strong, but O had widespread support (especially in the east), institutional presence

(majority in the Senate, control of crucial governorships and mayoralties) and cohesive organizations (PODEMOS).

Yet, I spent 2006 through 2008 disregarding this reduced institutional asymmetry. For instance, I insisted on approving articles by simple majority rather than supermajorities, hardly conceded any opposition demand, relied on street mobilizations to harass pro-O forces, held plebiscites while negotiations were still underway to flaunt the president’s national popularity, and moved the location of the meeting three times (first, to the suburb of Sucre, then to Lacua

Ñ, and finally to Oruro), to maximize O’s absence rates. At the Oruro meeting—a stronghold of pro-I support—a draft of the Constitution was approved without a single member of PODEMOS in attendance.

As in Venezuela, this pro-I de-alignment generated two predicable consequences. First, the constitutional draft approved in the city of Oruro (extreme table asymmetry, see Table 1), expanded presidential powers (see Graph 1). The second effect was to galvanize the

58 Gustafson 2006; Eaton 2007; Gamarra 2007; Gray-Molina 2008; Lehoucq 2008. 27 opposition. Essentially, O forces began to emulate some of I’s intransigent tactics. By 2007, O forces were also mobilizing social movements, inciting violence, carrying out unauthorized plebiscites (on departmental autonomy), calling for civil disobedience, seizing government buildings, and of course, rejecting every draft generated. By September 2008, Bolivia was close to the brink of war.

Only when I agreed to negotiate with O— i.e., to align table asymmetry with institutional asymmetry—was a truce possible. As in Nicaragua 1987, this agreement ultimately came about through both O and international pressure. Following a massacre between pro-O and pro-I social movements in the department of Pando in September 2008, the members of

UNASUR, a diplomatic block of Latin American countries, held a meeting to address the Bolivia issue. They offered full support for Morales, but demanded a compromise.

Morales budged. First, he agreed to move the negotiations to Congress, where O had a strong presence in the lower chamber and a majority in the Senate. He also agreed to create a special commission to revise the Oruro draft, comprised of an equal number of MAS and

PODEMOS representatives, plus seats for two other opposition parties. Table asymmetry thus became -20. Under this negative table asymmetry, presidentialism was lessened relative to the

Oruro draft (see Graph 1). Morales agreed to a series of concessions: not to seek re-election, abandoning the idea of reducing the size of the lower chamber and abolishing the Senate, accepting that future amendment would require two thirds of total votes in congress, rather than a simple majority. Moreover, 105 of the 411 articles of the Oruro draft were amended. A significant amendment involved expanding departmental autonomy. The constitutional draft 28 was approved by O delegates and by 61.4% of the electorate; civil strife subsided shortly thereafter.

In sum, de-alignment between table and institutional asymmetries influences not just constitutional texts, but also post-constitutional unrest. If the dealignment favors O, the constitution will lower presidential powers and create new post-constitutional opportunities for non-traditional forces to rise. If dealignment favors I, the result is hyperpresidentialism, O’s discontent, and possibly unrest.

VII. Splits among O: Ecuador 2008

Thus far, I have discussed the politics of constitutional reform as a bilateral bargain.

Often, however, O forces are split, converting the process into a trilateral affair. The split allows

I to negotiate with one sector of O to the detriment of the other. Ecuador 2008 illustrates this scenario.

In terms of power asymmetries, the Ecuador 08 case exhibits similarities and differences from the Venezuelan 1999 experience. As in Venezuela, pro-I asymmetry was large (the new president, Rafael Correa, won the presidency with a majority), and the president’s supporters were ideologically on the left, virulently anti-party , and very “trusting” of the presidency.59

59 See Freidenberg 2008. The anti-party sentiment was strong across the political system, not just within the Correa’s party. Only 33.7 percent of assembly delegates expressed agreement with the view that “democracy cannot exist without parties,” down from 85.7 among legislators in 1996. Furthermore, almost 80 percent of Correa’s delegates at the Constituent Assembly 29

However, there were two important differences. First, although Ecuador’s O forces were weak, de-institutionalized and volatile,60 they were not as weak as in Venezuela 1999, where the process of party collapse was more advanced. Ecuadorean parties still retained regional roots and could act as veto players.61 Second, there existed a more autonomous and vigorous sector of O: the non-traditional, more movement-oriented organizations, namely, the indigenous movements (CONAIE), the feminist movement, environmental organizations, and a variety of civic organizations. The difference with Venezuela is that these non-party forces did not become subordinate to I during the constituent assembly. 62

Essentially, I and the non-traditional, non-party O forces struck some sort of bargain.

The non-party O forces acceded to Correa’s plan to centralize a bit of power for the sake of weakening parties (whom these groups also viewed as political villains), but I had to agree to a series of concessions demanded by these forces (e.g., feminist groups obtained a multitude of rights).63 Nevertheless, I did not feel compelled to negotiate indefinitely, in part because had never belonged to a party, and 71.7 percent of expressed having “a lot of trust” on the presidency. See Freidenberg 2008.

60 Conaghan 2007.

61 Mejía Acosta, et al. 2008.

62 On how civil society groups gravitated toward Chávez in 1998-99, see Alvarez 2006.

63 in Ecuador as elsewhere in Latin America, these non-party organizations wanted to lower barriers to entry into the political world, even if that meant closing opportunities traditional parties. On this divide between parties and non-party representative organizations, see

Hochstetler and Friedman 2008; and Hellman 1992. 30

Correa’s political power expanded as the negotiations proceeded, based on his soaring approval rates (and declining approval rates for CONAIE). This expansion of power asymmetry encouraged Correa to end the negotiations abruptly: in the fall of 2008, he fired the Assembly’s president, Alberto Acosta, for yielding too much to social movements, and shortly thereafter finalized the draft.64

Ecuador 2008 can thus be explained by power asymmetries and splits among O.

Because I was stronger than any previous president in the last 20 years, the 2008 Constitution expanded presidential powers relative to the status quo. Because initially I needed to negotiate with the non-party side of O—which was not as weak as in Venezuela—the expansion of presidential powers did not match that of Venezuela 1999.

VIII. Negative Asymmetry and Aborted Rewrites

A more complete test of power asymmetry would consider the effects of the third variation of asymmetry: situations where O forces are stronger than I. For these cases, my argument predicts aborted constitutional change. Unable to muster the bargaining leverage to prevail in negotiations, the probability of ending in I’s rejection zone (right side of P I in Figure 1) increases . I thus has every incentive to use the power of office-holding to shelve the constitutional project entirely, or if O pressure is too strong, to consent only to a watered-down reform, such as changing just a few articles in the constitution (the amendment process), rather

64 Interview, Quito, January 2009. 31 than carrying out constitutional replacement. In short, presidents who are facing or anticipate declining powers will avoid constitutional rewrite.

Nine of the 10 cases in Table 1 partially confirm this hypothesis. In the majority of cases, I either enjoyed large power asymmetries prior to calling a constituent assembly

(Argentina, Brazil, Nicaragua, Paraguay, Colombia) or anticipated a large table asymmetry (e.g.,

Bolivia, Ecuador 08, Peru, Venezuela). The sole exception is Ecuador 98: I was an interim president with minimal powers, and the country was undergoing unrest. I launched a constituent assembly, despite his weakness, as a last-ditch effort to pacify the country.

Nevertheless, a more convincing approach would be to consider actual “negative” cases, i.e., those where major constitutional rewrite efforts were actually aborted by Is. 65 In the social sciences, it is difficult to identify non-occurring cases, but it is not impossible. One route is to follow Stokes 66 and study policy switches: individuals who make a promise as candidates but fail to deliver as presidents. In Latin America, several presidential candidates since the 1990s made constitutional rewrite a key campaign issue and faltered on their promises once elected.

While comparing platform and policy betrayal may never produce an exhaustive list of all possible negative cases, it can identify enough cases to do an initial probe of the negative asymmetry hypothesis.

Table 2 offers a list of cases of betrayed promises, drawing exclusively from the countries discussed in this paper. All Is in Table 2 campaigned strongly on behalf of

65 On the methodological merit of negative cases, see Ragin 2004

66 Stokes 2001. 32 constitutional change (source provided). In some cases, they even began formal preparations for major constitutional rewrite once in office. All of them, however, ultimately changed their mind, either avoiding constituent assemblies altogether or agreeing to small-scale amendments.

Table 2 also provides evidence of negative asymmetry for each case: I had a minority status in Congress, low approval ratings for I, outbreak of a political scandal involving I, splits and defections in the ruling party, or a combination of each. Negative asymmetry was a salient factor in all these cases of aborted constitutional change.

One could argue that an alternative explanation for the Chamorro, Samper, and

Pastrana cases might be “reform fatigue.” Having undergone constitutional change recently, these presidents might have sensed that the political system was too strained to undergo yet another constitutional overhaul. Perhaps. But in the cases of Caldera, Palacios, and Sánchez de

Lozada at least, reform fatigue cannot be the explanation since their immediate successors

(Chávez, Correa, and Morales) carried through on their promises to rewrite the constitution.

And in the case of Pastrana, his successor, Alvaro Uribe managed to get reelection regardless of levels of reform fatigue, in part because he became overwhelmingly popular and his opposition decomposed.

Finally, there are cases of presidents who did not campaign on behalf of constitutional change, but did spend energy in office orchestrating a major constitutional change, only to see their plans frustrated. This too was due to sudden negative political asymmetry. Carlos Menem in Argentina 1997-99 and Nicanor Duarte in Paraguay 2007, for instance, abandoned efforts to change the constitution to allow for reelection, due to splits in their parties and declining 33 approval ratings. Ernesto Pérez Balladares in Panamá 1999 lost a referendum to change the constitution to allow for his re-election, following a drop in approval ratings. Carlos Mesa in

Bolivia in 2005 resisted calls for a constituent assembly, and agreed only to watered-down constitutional reforms, mostly because he was an interim president associated with a collapsed administration. Manuel Zelaya in Honduras was removed from office in June 2009 for trying to go forward with an electoral “consultation” on whether to proceed with a referendum on whether to have a constituent assembly. Most analysts believed the President was seeking his reelection. Zelaya, however, faced growing negative asymmetry: his own party, the entire congress, the Supreme Court, and most media deemed the “electoral consultation” illegal (as well as other acts by the president). These powers use the military to remove the president from office, with few signs of popular disagreement.

In short, negative asymmetry is associated with aborted constitutional rewrites, either because I changes his or her mind in expectation of low bargaining leverage, or because O is strong enough to block Is efforts to use constitutional rewrite as a way to expand presidential powers. The evidence is not (and might never be) definitive, but it is plentiful.

VIII. Conclusion: Implications for Theories on Democratic Renewal

Constitutional rewrite occurs under many circumstances and cover many topics, as

Ackerman explained, but in democracies, the distance between I and O is crucial for understanding both the incidence of change and the resulting levels of presidential powers. In

Latin America since the 1980s, economic crises (in the form of runaway inflation, exchange rate 34 instability, and unsustainable debt), political crises (in the form of either regime transitions or political instability) and the expansion of civil society as a result of more open democratic systems raised the demand for new Constitutions, as is typical of worldwide cases of constitutional rewrite.67 I argued that the distance between I and O helps explain how this demand for change was met. If I felt far stronger than O (large asymmetry), it promoted constitutional replacement as a way to exploit its political power and secure more presidential powers.

Complications for I occurred if the distance between I and O was small or even negative.

Under low asymmetry—when I is stronger than O but only slightly—O could still pressure I to carry out a constitutional assembly with an outcome that was far more advantageous to O, and thus, more power diffusing. In this context, almost all forces gained something and all forces, especially I, ended up compromising. If, on the other hand, O was far stronger than I (negative asymmetry), the result was often aborted constitutional replacement. Aware of their declining bargaining leverage, Is simply avoided entering into negotiations, even if this meant betraying a campaign promise.

My selection of cases allowed me to control for a number of alternative explanations, including the idea that I’s ideology is an important determinant of constitutional outcomes.

Four presidents with similar ideologies (preference for centralism, economic statism, rhetorical embrace of “participatory” democracy, and disdain for traditional parties) yielded different constitutions; Nicaragua 87 reduced presidential powers, Ecuador 2008 expanded presidential powers but to a lesser degree than than Venezuela 1999, and Bolivia 07-09 expanded

67 See Elkins, Ginsburg, and Blount 2009. 35 presidential powers at first, generating much violence, until a new draft was renegotiated, which lowered presidential powers and violence. I’s ideology—similar across cases—cannot explain these various outcomes. My power asymmetry argument fared better.

The finding about the independent effect of power asymmetry has implications beyond the politics of constitutional making. The implications extend into the realm of the origins of self-enforcing democracy. Specifically, my findings depart from some well-established structuralist arguments about the proper route toward democratization. Structuralists tend to see democratization as contingent on maximizing the power of any given country’s democratizing class. They disagree on which group constitutes the key democratizing force: the modern bourgeoisie,68 organized labor 69 the middle classes,70 or the non-elite citizenry in general.71 Yet, they all agree that democratization requires a “transfer,” to use Acemoglu and

Robinson’s word,72 of formal powers from the traditional elite classes to the democratizing class and its allies. Once these democratizing classes get complete hold of the state (and displace old elites), democracy can happen.

My argument disagrees with both the protagonists and the storyline in these structural accounts. Total victory of one actor over another may not be the route to democratic renewal because, in constitution-making, it leads to an expansion of presidential powers relative to the

68 Moore 1966.

69 Collier 1999; Rueschmeyer et al. 1992

70 Luebbert 1991.

71 Tilly 1992; Acemoglu and Robinson 2006..

72 Acemoglu and Robinson 2006, 27. 36 status quo, which normally infuriates the opposition, and thus lessens the chance of loyalty. It also encourages I period to subsequently arrogate even more powers, which conspires against the notion of limited government, the sine qua non of democratic governance. A more reliable route to enduring democracy, this paper suggested, is to reduce rather than expand the power differential between political majorities and minorities at the moment of constitutional foundation. 37

I’s power

PI xI

PO

x SQ F O

O’s power Figure 1: Preferences for change in formal powers

Legend:

xI = I’s ideal point for changes in the constitution xO = O’s ideal point

Shaded area: agreement zone (win set).

SQ F = Status quo in terms of formal powers 38

I ’s power I’s power

xI

SQ PA

SQ PA xO

O’s power O’s power

Figure 2: Reduced Power Asymmetry Figure 3: High Pro-I Power Asymmetry

39

Graph 1:

40

Table 1: Institutional Asymmetries, Table Asymmetries and Changes in Presidential Powers

Country A B C D Institutional Table Asymmetry 74 Change in Hypothesis Asymmetry 73 Presidential Powers Confirmed

Colombia 75 20.00 (1990) -42.20 (1991) Reduction H2 Ecuador 97 76 -34.14 (1996) -40.00 (1997) Reduction H2 Argentina 77 -2.72 (1993) -24.20 (1994) Slight Reduction H3 Brazil 78 6.52 (1986) 6.00 (1988) Imperceptible ? Nicaragua 79 25.80 (1984) 9.00 (1987) Reduction H2 Peru 80 -31.20 (1990) 10.00 (1993) Expansion H1 Paraguay 81 33.34 (1989) 16.00 (1991) Reduction H2 Venezuela 82 -40.00 (1998) 86.26 (1999) Expansion H1/H4 Bolivia I (Sucre) 83 10.8 (2005) 7.4 (2006) No Agreement -- Bolivia II (Oruro) 84.9 (2007) Expansion H1/H4 Bolivia III (2009) -20 (2009) Reduction H2/H3 Ecuador 08 -100 (2006) 22.0 (2007) Expansion H1 Sources : (Political Database of the Americas 2007); for Colombia 1991: Cepeda; Ecuador 1996: (Payne, Zovatto et al. 2002); Ecuador 1997 :(CNN Election Watch 2007); Argentina 1993: (CDP 2007); Argentina 1994: (Tow 2007); Argentina 1995:(CNE 2007); Nicaragua 1984, 1987: (Reding 1987; Jonas and Stein 1990); Nicaragua 1990 : (CDP 2007; IDEA 2007); Paraguay 1989 : (Payne, Zovatto et al. 2002); Paraguay 1991 :(Facts on File World News Digest 1991; Payne, Zovatto et al. 2002) ; Venezuela 1998, 1999: (CDP 2007; CNE 2007); Bolivia 2005: (Europa

73 Difference in number of seats between I and O in the Congress (or in the lower Chamber if

Congress is bicameral). For Argentina 1994 and 1995, Colombia 1991, and Nicaragua 1984, percentage of vote, rather than seats, was used.

74 Difference in number of seats between I and O at the Constitutional Assembly (year)

75 I = Liberal Party 1990, 1991, 1994.

76 I = Social Christian Party 1996, 1997. Democracia Popular 1998.

77 I = Justicialista (or Peronist) Party. 1993, 1994, 1995.

78 I =Brazilian Democratic Movement Party, 1986, 1988,1990.

79 I = FSLN 1984, 1987; National Opposition Union 1990.

80 I = Fredemo 1990; New Majority Alliance – Cambio 90 1993, 1995.

81 I = National Republican Association 1989, 1991, 1993.

82 I =AD 1998; MVR-MAS-PPT 1999, 2000.

83 I =MAS 2005 and 2006. 41

World Year Book 2006); Bolivia 2006: (Latin American Regional Report: Andean Group 2006).

42

Graph 2:

Note: For Bol 07, the reported table asymmetry is the average between the Sucre meeting

(where negotiations began) and the Oruro meeting (where a draft was approved). 43

Table2: Negative Asymmetry and Non-Occurring Constituent Assemblies

Country I Evidence that I Factors leading to Outcome campaigned on negative asymmetry behalf of constitutional rewrite Bolivia Jaime Paz Zamora Van Cott 2005 Minority President No Constitutional assembly Gonzalo Sánchez de Van Cott 2000; Minority President No constitutional Lozada 2005 assembly; reform by legislative negotiation Colombia Ernesto Samper Restrepo M. 1996 Corruption scandal; No constitutional defections from ruling assembly party Andrés Pastrana Cepeda Ulloa 2006. Minority President No Constitutional assembly Ecuador Alfredo Palacios Latin American Interim president; low No constitutional Regional Report: approval ratings; change Andean Group impeachment threats 2005:6. Nicaragua Violeta Chamorro McConnell 1997; Divisions within the Watered-down Esgueva Gómez ruling coalition amendment 2003 rather than a Constituent Assembly Venezuela Rafael Caldera Alvarez 1998; Minority president; low No Constitutional Combellas 1998. approval ratings Change

44

Appendix 1: List of presidential powers

Source: author, unless otherwise specified

Categories and Scores Descriptions of Sub-Categories Presidential Legislative Powers (36)

Package Veto/Override 4= Veto with no override 3= Veto with override requiring 2/3 majority 2 = Veto with override requiring between simple and 2/3 majority 1 = Veto with override requiring simple majority 0 = No veto (Shugart and Carey 1992 modified by author)

Partial Veto/Override 4= Veto with no override 3= Veto with override requiring 2/3 majority 2 = Veto with override requiring between simple and 2/3 majority 1 = Veto with override requiring simple majority 0 = No veto (Shugart and Carey 1992 modified by author

Decree 4 = Reserved powers, no rescission 3= President has temporary decree authority that includes regulation of constitutional amendments 2 = President has temporary decree authority / authority to enact decrees limited 0 = No decree powers, or only as delegated by assembly (Shugart and Carey 1992 modified by author)

Exclusive Introduction of 4 = Restricted amendment of introduced legislation by assembly Legislation (Reserve Policy Areas) 2 = Unrestricted amendment of introduced legislation by assembly 0 = No exclusive introductory powers (Shugart and Carey 1992 modified by author). 45

Budgetary Powers 4 = President prepares budget, no amendment permitted 3 = Assembly may reduce but not increase the amount of budgetary items 2 = President sets upper limit on total spending, within which assembly can amend 1 = Assembly may increase expenditures only if it designates new revenues 0 = Unrestricted authority of assembly to prepare or amend budget (Shugart and Carey 1992).

Recall Referendum for Legislators 4= Recall referendum with low threshold for removal 2= Recall referendum with high threshold for removal 0= No recall referendum

Treaty Power 4= President may enter into treaty's without congressional approval in all areas 2= President may enter into treaty's without congressional approval in certain areas 0= President cannot enter into treaty's without congressional approval

Presidential Recess Powers 4= President can carry out some responsibilities of the legislature unilaterally during a recess 2= President can carry out some responsibilities of the legislature, but only in concert with members of legislature during a recess (usually through a permanent committee) 0= President cannot carry out any responsibilies of the legislature during a recess

Enabling Law 4= Enabling laws explicitly possible in every area 2.66= Enabling laws explicitly possible in many areas 1.33= Enabling laws explicitly possible in few areas 0= No explicit enabling law mechanism

Presidential Non-Legislative Powers (40) Cabinet Formation 4 = President appoints cabinet without need for confirmation or investiture 2.66 = President names cabinet members subject to confirmation or investiture by assembly 1.33 = President names premier, subject to investiture, who then names other ministers 0 = President cannot name ministers except upon recommendation by the assembly (Shugart and Carey 1992). 46

Censure of the President 4= No censure possible 2= Censure with high threshold (i.e. 2/3 majority) 0= Censure with low threshold (i.e. simple majority)

Censure of Cabinet and Ministers 4 = Assembly may not censure and remove cabinet or ministers 2 = Censure with high threshold (i.e. 2/3 majority) 0 = Censure with low threshold (i.e. simple majority) (Shugart and Carey 1992 modified by author

Cabinet Dismissal 4 = President can dismisses cabinet members at will 2 = Powers of dismissal restricted by the legislature 0 = Cabinet or ministers may be removed only by assembly (Shugart and Carey 1992 modified by author

Dissolution of Assembly 4 =Uunrestricted 3 = Restricted by frequency or point within term 2 = Restricted, only as a response to censures 1 = Requires new presidential election 0 = No provision (Shugart and Carey 1992 modified by author)

Preventative Detention (outside of 4= Unlimited (or de facto unlimited) detention powers states of exception) 2= Limited detention powers 0= No powers of preventative detention

Presidential Reelection 4 = No term limits 3 = Consecutive reelection; third reelection with restrictions (sitting out, only one more time) 2 = Consecutive reelection 1 = One term, re-election possible after sitting out 0 = No reelection

47

Presidential Term Length 4= 7 years 2.66= 6 years 1.33= 5 years 0= 4 years Shugart and Carey 1992 modified by author)

Recall Referendum for the 4= No recall referendum President 2= Recall referendum possible with high threshold for removal 0= Recall referendum possible with low threshold for removal

Appointment/Election of Judges 4= President may appoint judges without legislative oversight/approval 2= President may appoint judges, but only with legislative oversight/approval 0= Legislature appoints judges / people elect judges

Presidential Procedural Powers (24)

Presidential Initiation of 4= President can initiate a constitutional amendment, but requires Constitutional Amendments legislature/referendum to pass 0= President cannot initiate constitutional amendments

Presidential Initiation of 4= President can initiate a constitutional rewrite, but requires Constitutional Rewrites legislature/referendum to pass 0= President cannot initiate constitutional rewrites

Military Appointments and 4= No legislative approval required Promotions 2= Legislative approval only for rank of general officer/admiral or above 0= Legislative approval required (Author); Idea from (Bejarano 2006) 48

Presidential Immunity to 4= Unrevokable presidential immunity Prosecution 2= Presidential immunity revokable by another branch of the government 0= No presidential immunity to prosecution

Proposal of Legislative Referenda 4 = Unrestricted 2 = Restricted 0 = No presidential authority to propose referenda

Power to Declare a State of 4= Unlimited power of the president to declare a state of exception/emergency Exception/Emergency 2= Power to declare state if exception/emergency subject to congressional/judicial revocation/approval 0= No presidential power to declare states of exception/emergency (Bejarano 2006 modified by author)

Degree of Presidential Indpendence from the Legislature (20) Run-Off Selection 4= No power to select the president in a run-off 0= Power to select the president in a run-off under certain conditions

Amendment Capabilities of the 4= No amendement powers for legislation Legislature for Legislation 2= Limited amendments for legislation 0= Unlmited amendments for legislation

Impeachment of President by 4= No impeachment or inter-branch mechanism required Legislature 2= High threshold for impeachment 0= Low threshold for impeachment

49

Immunity of Legislators to 4= No legislator immunity Prosecution 0= Some legislator immunity

Investigative Powers of the 4= No ability to initiate investigations of the executive branch ministers Legislature over the Executive 2= Ability to initiate but not run investigations of the executive branch ministers 0= Ability of congress to initiate and run investigations of the executive branch ministers

Legislative Camerality

Degree of Presidential Independence from the Judiciary (16) Constitutional / Supreme Court 4= No provision for top court 3= Primarily dependent on president 2= Mixed dependency, dismissal determined by law or uncertain dependency 1= Dependent on assembly 0= Life appoitment of justices (life appointment constrains power of subsequent presidents) (Alberts 2006 modified by author)

Judicial Review Powers Available to 4= No judicial review powers granted to the judiciary the Judiciary 2= Judicial review powers limited to the supreme/constitutional court 0= Judicial review powers at most levels of the federal judiciary

50

Independent Impeachment of 4= No judicial impeachment of president President by Judiciary 0= High court may impeach the president

Prosecutor General 4= No provision for the prosecutor general 3= Dependent on president 2= Mixed dependency (requires non-presidential approval) 1= Determined by civil society or professional groups 0= Dependent on the assembly (Alberts 2006 modified by author)

Degree of Presidential Power over the Federal System

Local Elections 4= No election of sub-national officials 3= Elections for mayors (not for governors) 2= Elections for governors (not for mayors) 1= No direct elections for executive authority in the capital city or some other major jurisdiction (elections for all other governors and mayors) 0= Elections for all governors and mayors

51

References

Acemoglu, Daren and James A. Robinson. 2006. Economic Origins of Dictatorship and Democracy . New York: Cambridge University Press. Alberts, Susan. 2006. "Subjecting Power to Rules: Constitutionalism and Democratic Survival in Latin America." American Political Science Association . Philadelphia, PA. Alston, Lee J., Marco Melo, Bernardo Mueller and Carlos Pereira. 2005. "Political Institutions, Policymaking Processes and Policy Outcomes in Brazil." Latin American Research Network Working Paper No. R-509 . Washington, DC, Inter- American Development Bank. Alvarez Díaz, Angel. 1999. Revista SIC . Alvarez, Tulio A. 1998. La constituyente: Todo lo que usted necesita saber [The Constituent Assembly: Everything You Need to Know] . Caracas: Editorial CEC, Libros de El Nacional. Arato, Andrew. 2005. "Constitutional Learning." Theoria . Archer, Ronald P. and Matthew Soberg Shugart. 1997. "The Unrealized Potential of Presdential Dominance in Colombia." In Scott Mainwaring and Matthew Soberg Shugart, eds., Presidentialism and Democracy in Latin America . New York: Cambridge University Press. Armony, Ariel C. and Héctor E. Schamis. 2005. "Babel in Democratization Studies." Journal of Democracy 16(4): 116-117. Azicri, Max, Ed. 1991. Examining Some Relevant Political and Constitutional Issues of the 1987 Nicaraguan Constitution , The Nicaraguan Constitution of 1987, Athens, OH, Ohio University Monographs in International Studies. Bejarano, Ana María. 2006. "Placing the Constitution of 1991 in Time: Colombia’s Long Road to Democratization." Presented to the APSA Convention, Philadelphia, Aug. 30-Sept. 2. Bermeo, Nancy. 1997. "Myths of Moderation: Confrontation and Conflict during Democratic Transitions." Comparative Politics 29: 305-322. Blanco, Carlos. 2003. Revolución y desilución. La Venezuela de Hugo Chávez [Revolution and Disappointment: Hugo Chávez’s Venezuela]. Madrid: Catarata. Calvert, Peter. 2002. "Argentina: The Crisis of Confidence." Annual Conference of the UK Political Science Association. Cárdenas, Mauricio, Roberto Junguito and Mónica Pachón. 2004. "Political Institutions and Policy Outcomes in Colombia: The Effects of the 1991 Constitution." Washington, D.C., Inter-American Development Bank, Research Department. CDP. 2007. "Argentina 1993 Legislative Elections." Center on Democratic Performance, Electoral Results Archive. ———2007. "Nicaragua 1990 Legislative Elections." Center on Democratic Performance, Elections Results Archive. ———2007. "Venezuela 1998 Legislative Elections." Center on Democratic Performance: Election Results Archive. 52

Chávez, Hugo, Martha Harnecker and Chesa Boudin. 2005. Understanding the Venezuelan Revolution: Hugo Chavez Talks to Marta Harnecker . New York: Monthly Review Press. CNE. 2007. "Elecciones, República Bolivariana de Venezuela." Consejo Nacional Electoral. ———2007. "Resultados de Elecciones Presidenciales Anteriores." CNN Election Watch. 2007. "Ecuador (National Assembly)." cnn.com. Collier, Ruth Berins. 1999. Paths Toward Democracy: the Working Class and Elites in Western Europe and South America . New York: Cambridge University Press. Colomer, Josep Maria. 1995. "Strategies and Outcomes in Eastern Europe." Journal of Democracy 6: 74-85. Combellas, Ricardo. 1998. "Introducción." Constituyente: Aportes al debate , ed. Ricardo Combellas, Caracas: ed. Fundación Honrad Adenauer, COPRE : xi. ———2003. "El proceso Constituyente y la Constitución de 1999." Caracas : 16. Conaghan, Catherine M. 2007. "Notes on Recent Elections: The 2006 Presidential and Congressional Elections in Ecuador." Electoral Studies 2. Coppedge, Michael. 2003. "Venezuela: Popular Sovereignty versus Liberal Democracy." In Jorge Domínguez and Michael Shifter, eds., Constructing Democratic Governance in Latin America . Baltimore, MD: Johns Hopkins University Press. Corrales, Javier and Michael Penfold. 2007. "Venezuela: Crowding Out the Opposition." Journal of Democracy . Crisp, Brian and Gregg Johnson. 2001. "De instituciones que restringen a instituciones ausentes" [From restricting institutions to missing institutions] In José Vicente Carrasquero, Thaís Maingón and Friedrich Welsch, eds., Venezuela en transición: elecciones y democracia, 1998-2000 Caracas: Red Universitaria de Estudios Políticos de Venezuela, CDB Publicaciones. Di Palma, Giuseppe. 1990. To Craft Democracies: an essay on democratic transitions . Berkeley, CA: University of California Press. Eaton, Kent. 2007. "Backlash in Bolivia: Regional Autonomy as a Reaction against Indegenous Mobilization." Politics and Society 35(1): 71-102. Elkins, Zachary, Tom Ginsburg, and Justin Blount. 2009. Can We Trust Legislators to Write Constitutions? Paper presented at the Latin American Studies Association Congress, Rio de Janeiro. Escovar Fornos, Iván. 2000. El constitucionalismo nicaragüense, vol. 2 .[Nicarauguan constitutionalism]. Managua: Editorial Hispamer. Esgueva Gómez, Antonio. 2003. Historia Constitucional de Nicaragua [Constitutional History of Nicaragua]. Managua: Lea Grupo Editorial. Europa World Year Book. 2006. "Bolivia." London, Europa Publications Limited. Facts on File World News Digest. 1991. "Constitutional Assembly Elected: Paraguay." Facts on File. Finer, S. E., Vernon Bogdanor and Bernard Rugden. 1995. Comparing Constitutions. Oxford University Press. Freidenberg, Flavia. 2008. "¿Renovación o continuismo? Actitudes, valores y trayectoria de la clase política ecuatoriana." [Renewal or continuity? Attitudes, 53

values and career paths across Ecuador’s political class]. Salamanca, Instituto de Iberoamérica, Universidad de Salamanca. Gamarra, Eduardo A. 2007. "Bolivia on the Brink." New York, NY, Council on Foreign Relations, Center for Prevention Action. Geddes, Barbara. 1994. Politicians' Dilemma . Berkeley, CA: University of California Press. Ginsburg, Tom, Zachary Elkins and James Melton. 2007. "The Lifespan of Written Constitutions." American Law and Economics Association Annual Meetings , The Berkeley Electronic Press. Gray-Molina, George. 2008. "The Post-Neoliberal Challenge in Bolivia: Any Room for the Left in the Global Economy?" What Does the Left Do Right? Austin, TX. Gustafson, Bret. 2006. "Spectacles of Autonomy and Crisis: Or, What Bulls and Beauty Queens Have To Do with Regionalism in Eastern Bolivia." Journal of Latin American Anthropology 11(2): 351-379. Hartlyn, Jonathan and Juan Pablo Luna. 2009. "Constitutional Reform in Latin America: A Framework for Analysis." Paper presented at the Latin American Studies Association Congress, Rio, Brazil. Hellman, Judith Adler. 1992. "The Study of New Social Movements in Latin America and the Question of Autonomy." In Arturo Escobar and Sonia E. Alvarez, Boulder, eds., The Making of Social Movements in Latin America . CO: Westview Press. Higley, John and Richard Gunther. 1992. Elites and Democratic Consolidation in Latin America and Southern Europe. Cambridge University Press. Hochstetler, Kathryn and Elisabeth Jay Friedman. 2008. "Can Civil Society Organizations Solve the Crisis of Partisan Representation in Latin America?" Latin American Politics and Society 50(2): 1-26. Horowitz, Donald. 2002. "Constitutional Design: Proposals versus Processes." In Andrew Reynolds, ed., The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy . Oxford University Press : 27, 35. Huntington, Samuel. 1991. The Third Wave: Democratization in the Late Twentieth Century . Norman, OK: University of Oklahoma Press. IDEA. 2007. "Voter Turnout: Nicaragua." International Institute for Democracy and Electoral Assistance. Jonas, Susanne and Nancy Stein. 1990. "The Construction of Democracy in Nicaragua." Latin American Perspectives 17(3): 10-37. Jones, Mark P. 1997. "Evaluating Argentina´s Presidential Democracy." In Scott Mainwaring and Matthew Soberg Shugart, eds., Presidentialism and Democracy in Latin America . Cambridge University Press, 290-298. Kantor, Harry. 1977. "Efforts Made by Various Latin American Countries to Limit the Power of the President." In Thomas V. DiBacco, ed., Presidential Power in Latin American Politics . New York: Praeger. Karl, Terry Lynn. 1990. "Dilemmas of Democratization in Latin America." Comparative Politics 23: 1-23. 54

Karl, Terry Lynn and Phillipe Schmitter. 1991. "Modes of Transition in Southern and Eastern Europe and South and Central America." International Social Science Journal 128: 269-284. Kay, Bruce H. 1996. "`Fujipopulism´ and the Liberal State in Peru, 1990-1995." Journal of Interamerican Studies and World Affairs 38(4). Latin American Regional Report: Andean Group. 2005. Latin American Regional Report: Andean Group : 6. ———2006. "Rows as Constituent Assembly Looms." Latinnews. Lehoucq, Fabrice. 2008. "Bolivia’s Constitutional Breakdown." Journal of Democracy 19(4). Lijphart, Arend. 1991. "Constitutional Choices for New Democracies." Journal of Democracy 2(1): 72-84. Linz, Juan J. 1978. The Breakdown of Democratic Regimes . Baltimore, MD: Johns Hopkins University Press. Llanos, Mariana. 2003. Privatization and Democracy in Argentina: An Analysis of President-Congress Relations . Hampshire, Great Britain: Palgrave. Luebbert, Gregory M. 1991. Liberalism, Fascism, or Social Democracy: Social Classes and the Political Origins of Regimes . New York: Oxford University Press. Lutz, Donald S. 2006. Principles of Constitutional Design . New York, NY: Cambridge University Press. Maingón, Thaís, Carmen Pérez Baralt and Heinz R. Sonntag. 2000. "La batalla por una nueva Constitución para Venezuela." Revista Mexicana de Sociología 62: 91- 124. Márquez Ferrer, Victorino. 1998. "Verdades y mentiras de la constituyente" [Truths and Lies surrounding the Constituent Assembly] In Ricardo Combellas, ed., Constituyente: Aportes al debate [Constituent Assembly: Contributions to the Debate]. Caracas: Fundación Honrad Adenauer, COPRE. McConnell, Shelley. 1997. "Institutional Development." In Thomas W. Walker, ed., Nicaragua without Illusions: Regime Transition and Structural Adjustment in the 1990s . Wilmington, DE: ed. SR Books. Mejía Acosta, Andrés, María Caridad Araujo, Aníbal Pérez-Liñán and Sebastián Saiegh. “Veto Players, Fickle Institutions, and Low-Quality Policies: The Policymaking Process in Ecuador.” Policymaking in Latin America: How Politics Shapes Policies , ed. Ernesto Stein, Mariano Tommasi, Carlos Scartascini and Pablo Spiller. Cambridge: Harvard University Press. Mijeski, Kenneth J. 1991. "Introduction." In Kenneth J. Mijeski, ed., The Nicaraguan Constitution of 1987 . Athens, OH: Ohio University Center for International Studies. Moore, Barrington. 1966. Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World . Boston: Beacon Press. Negretto, Gabriel. 1998. "Constitution-Making and Institutional Design: The Reform of Presidentialism in the Argentine Constitution of 1994." LASA . Chicago. ———2001. "Negociando los poderes del presidente: reforma y cambio constitucional en la Argentina." Desarrollo Económico 163(41). 55

———2002. "Gobierna Solo el Presidente? Poderes de Decreto y Diseño Institucional en Brasil y Argentina." Desarrollo Económico 167(42). ———2009. "Political Parties and Institutional Design: Explaining Constitutional Choice in Latin America." British Journal of Political Science 39: 117-139. Nolte, Detlef. 2009. "Constitutional Change in Latin America from a Comparative Perspective." Paper presented at the Meeting of the Latin American Studies Association, Rio de Janeiro, Brazil, June. North, Douglas C., William Summerhill and Barry R. Weingast. 2000. "Order, Dirsorder, and Economic Change: Latin America versus North America." In Bruce Bueno de Mesquita and Hilton L. Root, eds., Governing for Prosperity . New Haven, CT: Yale University Press : 23-29. O’Donnell, Guillermo. 1994. "Delegative Democracy." Journal of Democracy 5(1): 55- 69. O'Donnell, Guillermo and Phillipe Schmitter. 1986. Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies . Baltimore, MD: Johns Hopkins University Press. Olson, Mancur. 2000. Power and Prosperity . New York, NY: Basic Books. Pastor, Robert A. 2002. Not Condemned to Repetition: The United States and Nicaragua, 2nd Edition . Boulder, CO: Westview Press. Payne, J. Mark, Daniel G. Zovatto, Ferando Carrillo Flórez and Andrés Allamand Zavala. 2002. Democracies in Development: Politics and Reform in Latin America . Washington, D.C.: Inter-American Development Bank. Penfold, Michael. 1999. "Constituent Assembly in Venezuela: First Report." Atlanta, GA, Carter Center. Persson, Torsten and Guido Tabellini. 2003. Economic Effects of Constitutions . Cambridge, MA: The MIT Press. Political Database of the Americas. 2007. "Electoral Data." Georgetown University and Organization of the American States. Prevost, Gary. 1997. "The FSLN." In Thomas W. Walker, ed., Nicaragua without Illusions: Regime Transition and Structural Adjustment in the 1990s . Wilmington, DE: SR Books. Przeworski, Adam. 1991. Democracy and the market: political and economic reforms in Eastern Europe and Latin America . New York, NY: Cambridge University Press. Ragin, Charles C. 2004. "Turning the Tables: How Case-Oriented Research Challenges Variable-Oriented Research." In Henry E. Brady and David Collier, eds., Rethinking Social Inquiry: Diverse Tools, Shared Standards . Lanham, MD: Rowman and Littlefield Publishers. Reding, Andrew. 1987. "Nicaragua's New Constitution: A Close Reading." World Policy Journal 4(2). Restrepo M., Luis Alberto. 1996. "El Ejecutivo en la crisis: dimensiones, antecedentes y perspectivas." In Francisco Leal Buitrago, ed., Tras las huellas de la crisis Política . Bogotá: Tercer Mundo Editores/Fescol/IEPRI. Reynolds, Andrew, Ed. 2002. The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy Oxford University Press. 56

Rueschemeyer, Dietrich, Evelyne Huber and John D Stephens. 1992. Capitalist Development and Democracy . Chicago: Chicago University Press. Schedler, Andreas, Mark F. Plattner and Larry Diamond, Eds. 1999. The Self Restraining State: Power and Accountability in New Democracies. Lynne Reinner Publishers, Inc. Schmidt, Gregory D. 2000. "Delegative Democracy in Peru? Fujimori's 1995 Landslide and the Prospects for 2000." Journal of Interamerican Studies and World Affairs . Shugart, Matthew Soberg and John M. Carey. 1992. Presidents and Assemblies: Constitutional Design and Electoral Dynamics. Cambridge University Press. Snyder, Richard and James Mahoney. 1999. "The Missing Variable: Institutions and the Study of Regime Change: Review Article." Comparative Politics 32(1): 103- 122. Stokes, Susan. 2001. Mandates and Democracy: Neoliberalism by Surprise . New York, NY: Cambridge University Press. Tilly, Charles. 1992. Coercion, Capital, and European States, AD 990-1992. Cambridge, MA: Blackwell ———1998. "Where do Rights Come From." In Theda Skocpol, ed., Democracy, Revolution and History . Cornell University Press. United States Department of State. 1991 [1987]. "The Sandinista Constitution." In Kenneth J. Mijeski, ed., The Nicaraguan Constitution of 1987 . Athens, OH: Ohio University Monographs in International Studies. Viciano Pastor, Roberto and Rubén Martínez Dalmau. 2001. Cambio político y proceso constituyente en Venezuela, 1998-2000 [Political Change and Constituent Process in Venezuela] . Caracas: Vadell Hermanos. Weingast, Barry R. 2004. "Constructing Self-Enforcing Democracy in Spain." In Erwin L. Morris, Joe A. Oppenheimer and Karol Edward Soltan, eds., Politics from Anarchy to Democracy: Rational Choice in Political Science . Stanford, CA: Stanford University Press. Wilpert, Gregory. 2003. Venezuela's New Constitution , at http://www.venezuelanalisis.com . World Bank. 2002. Globalization, Growth and Poverty . Washington, DC: The World Bank.